The North Holland Dune Reserve stretches roughly from 52.487°N, 4.590°E to 52.682°N, 4.691°E and has an area of roughly 53 km². We did most of our observations in the area just south of the village Bergen aan Zee at two main localities. One is a small patch of woodland surrounded by an open dune area relatively close to the inland border of the North Holland Dune Reserve, with English Oak (Quercus robur L., Fagaceae) as the main tree species and F. vesca dominant in the herb layer, further indicated as ‘L1_Oak’. The other locality is a small patch of woodland surrounded by an open dune area relatively close to the North Sea with Birch (Betula sp., Betulaceae) as the main tree species and again F. vesca dominant in the herb layer, further indicated as ‘L2_Birch’. In addition, we carried out some observations in the surrounding area. An overview of all localities and dates is provided in Table 1. JvR took the photographs, with a Canon EOS 7D camera with a Canon EF 100mm f/2.8L Macro IS USM lens, except for the photographs in figures 15 and 16, which were taken with a Nikon D80 camera with a Micro-Nikkor AF 60mm f/2.8 D lens.

On 29.x.2013 and 19.xi.2013, we collected spatulate cases on Fragaria vesca at the L2_Birch locality. We kept the collected larval cases in a small transparent plastic jar and put them in a garden shed to hibernate, along with small pieces of the host plant. In March 2014 we transferred the cases to fresh Fragaria leaves in the garden. We first observed fresh feeding traces on 11.iv.2014, the active larvae still with spatulate cases. On 05.v.2014 we observed that the most active larva had stopped feeding and that the case had moved to the petiole of the leaf it had been feeding on. The case now had a trivalved appearance. We transferred the case into a small plastic jar and kept it indoors until emergence. We found the second case lying on a leaf rather than being attached to the underside and we assumed that the larva had died. We obtained another rearing result from a still fresh spatulate case on F. vesca on 07.v.2014 in the dunes near Egmond (GPS coordinates 52.637°N, 4.649°E) by collecting the case together with fresh leaves that were taken indoors until emergence.

We selected ten specimens from different localities and hosts for DNA analysis, see Table 2. We pulled the larvae from their case with forceps, damaging the case as little as possible. The cases are stored as vouchers in the RMNH collection. For DNA analysis we used the mitochondrial COI-barcode gene region (Hebert et al. 2003). DNA extraction and PCR amplification followed the methods described in van Nieukerken et al. (2012). We added the collecting and sequence data of the specimens to the Barcode of Life Datasystems (BOLD; Ratnashingham and Hebert 2007), under their unique RMNH registration number as sample ID. Sufficient reference material of all potential Coleophora species that feed on Fragaria was available in BOLD and we made identifications of the larvae by examining the distance to the nearest neighbour. There was a 100% match for most samples, and also the monophyly criterion was met, when we evaluated whether our sequence fitted within a monophyletic DNA barcode species cluster.

RMNH Rijks Museum voor Natuurlijke Historie collection, housed at Naturalis Biodiversity Center, Leiden, The Netherlands.

ZMA Zoölogisch Museum Amsterdam collection, housed at Naturalis Biodiversity Center, Leiden, The Netherlands.

In the spring of 2011 at locality ‘L1_Oak’, during an intensive search for Tinagma perdicella Zeller, 1839 (Douglasiidae), which flies during daytime and can be found close to Fragaria, JvR noticed occupied Coleophoridae larval cases and feeding damage on Fragaria (Figs 1, 2). These spatulate cases at first impression had a shape similar to cases of Coleophora albicostella, and given the host plant, led to the assumption that this was the rare C. albicostella. However, rearing was unsuccessful, leaving the identifications unconfirmed. In the autumn of that same year, again spatulate cases were found, this time at locality ‘L2_Birch’. At that time rearing was not attempted. It was assumed to be easier to look for larval cases in the spring and attempt to rear those, allowing the larvae to hibernate under natural conditions. However, in the spring of 2012 JvR encountered no cases and only in the next autumn, JvR observed cases again.

Figures 1–8. 

1. Spatulate case found on Fragaria 20.v.2011. To the right of the case feeding damage can be seen. 2. Spatulate case found on Fragaria 20.v.2011, with larva. 3. First stage case on Fragaria, showing also the place where the leaf was cut for a second stage case, 02.x.2012. 4. Spatulate case on Fragaria, 02.x.2012. 5. Fragaria leaf from Figure 3 showing feeding damage, first stage case and ‘leaf cut’, 02.x.2012. 6. First stage cases and spatulate cases on Rosa, 02.x.2012. 7. Tiny fleck mines on Fragaria, 29.x.2013. 8. Small pistol case on Fragaria, 29.x.2013.

Figures 9–16. 

9. Spatulate leaf case on Fragaria collected for DNA analysis, 29.x.2013. 10. Spatulate leaf case on Rosa collected for DNA analysis, 29.x.2013. 11. Final, trivalved case of the reared larva after hibernation and after feeding ended on 05.v.2014. 12. Fleck mines produced by the reared larva after hibernation and after feeding ended on 05.v.2014. 13. Emerged imago Coleophora gryphipennella reared from a hibernating larva. 14. Emerged imago Coleophora gryphipennella reared from a hibernating larva. 15. Fresh green case on Fragaria, 07.v.2014, leg. and photograph Luc Knijnsberg. 16. Emerged imago from the case found on 07.v.2014, photograph Luc Knijnsberg.

In the autumn of 2012 at locality ‘L2_Birch’ we stumbled upon a small Rosa sp. bush with almost each leaf occupied by a Coleophora case, and most leaves were eaten out. At the same locality, a few meters further, we discovered a Fragaria vesca plant also with typical Coleophora feeding traces and with cases indistinguishable from the ones on the Rosa sp. bush. Both even showed the small, approximately three mm long, rectangular shaped first stage cases still present on the plants (Figs 3–6). Only three cases were found on F. vesca and JvR decided to return in the spring to collect cases rather than to attempt to hibernate and rear these cases.

However, in the spring of 2013 again JvR found no larval cases on Fragaria. It took until October 2013 until finally JvR could collect cases. At this point collaboration started with CD to use DNA barcode analysis to identify the species that creates these cases on Fragaria and Rosa. To collect fresh material, we intensified the search for larval cases and this yielded two cases from Rosa spp. as well as from Fragaria vesca from different localities (Table 1). During this collecting effort, we found a second type of feeding trace and larval Coleophora case on F. vesca: tiny fleck mines created by larvae in small pistol-shaped cases (Figs 7–8) at locality ‘L1_Oak’. We sequenced a total of ten larvae (see Figs 8–10) from the different localities on 19.x.2013 and 29.x.2013 (Table 2).

Nine out of the ten barcoded specimens yielded a DNA barcode. All identifications were unambiguous (Table 2). None of the specimens proved to be Coleophora albicostella. Instead, a mixture of three other species, none of which had been recorded to feed on Fragaria, was found: C. gryphipennella (Hübner, 1796), C. lutipennella (Zeller, 1838), and C. flavipennella (Duponchel, 1843). The larger cases, i.e. >5 mm long, were all of the spatulate type and were identified as C. gryphipennella. The smaller cases were pistol shaped. These were either C. lutipennella or C. flavipennella, species that normally feed on oaks (Quercus spp.). We did not find larger pistol shaped cases.

We collected a few autumn cases from Fragaria vesca in November 2013 for rearing. They hibernated in a garden shed and began feeding on Fragaria again in the spring (Figs 11–12). On 21.v.2014 an adult emerged. Based on the external characters we identified it as Coleophora gryphipennella, which can be readily distinguished from C. albicostella by the lack of a white stripe along the costa (see Figs 13–14). In addition, a fresh case collected by Luc Knijnsberg on F. vesca on 7.v.2014 (Fig. 15) at ‘L1_Oak’ also successfully yielded an imago of C. gryphipennella on 10.vi.2014 (Fig. 16).

We initially expected to have found the rare Coleophora albicostella in the North Holland Dune Reserve, but none of the specimens that we reared or analysed for DNA proved to be C. albicostella. Although this does not completely exclude the option that C. albicostella occurs on Fragaria vesca in the coastal dunes of The Netherlands, further investigation of the host records of this species made it even more unlikely. Fragaria spp. is first mentioned as a host plant for C. albicostella by Hering (1957), along with Potentilla spp. and Rubus spp. The status and source of these observations are unclear, as they are mentioned without reference to the stage (e.g. larval or adult stage) of the studied material or whether rearing was attempted. Hering (1957) does provide a drawing of the larval case, which appears as a rather general looking spatulate case that cannot be distinguished from, for example, C. gryphipennella. Furthermore, the case may include slight differences based on the host plant that was included in its construction. Huemer (1988) only mentions Potentilla spp. and occasionally also Comarum spp., Filipendula spp., Rubus spp., and Sanguisorba spp. as hosts (all Rosaceae). Rearing has been documented from Potentilla spp. (Bryner 2015; Richter 2015) and Geum spp. (Hugo van der Wolf, pers. comm., in collection). We believe that these records represent the predominant hosts. Even if C. albicostella may feed on the other genera, it is likely only so in areas where Potentilla or Geum may be found. In any case, also in the light of our findings here, C. albicostella cannot readily be reported when spatulate larval cases are encountered on Fragaria and should always be reared, or DNA barcoded, to efficiently identify such larvae.

We demonstrated both by DNA analysis on the larvae from spatulate shaped larval cases collected on Fragaria vesca in the fall as well as by successfully rearing adults from larvae that have been feeding on F. vesca both in the fall and in the spring that F. vesca is a novel host for Coleophora gryphipennella. At least in the study area, this seems unlikely to be due to a shortage of Rosa, especially Rosa pimpinellifolia L. but also larger species such as Rosa rubiginosa L. are abundant and C. gryphipennella can be found in many localities feeding on Rosa throughout the habitat. The adaptation to another host plant for C. gryphipennella could be a local habit, only occurring in the dunes of the North Holland Dune Reserve. On the other hand, it could also be a more general, widespread habit that can be found throughout its distribution in Europe. More evidence to support this came from a recent find of a C. gryphipennella larva on F. vesca near Durbuy, Belgium (Steve Wullaert, pers. comm.), as confirmed by DNA analysis (RMNH.INS.30422 on BOLD). C. gryphipennella is distributed throughout most of Europe, Turkey, and Central and Eastern Siberia (Baldizzone et al. 2006). C. gryphipennella is mostly described as monophagous on Rosa spp. Our observations include feeding signs and larval cases on Rosa pimpinellifolia and Rosa rubiginosa. Rubus sp. and Rubus corylifolius Sm. (Rosaceae) are mentioned as incidental host plants by Hering (1957). Hering describes that in the fall the cases are somewhat pressed together sideways and are terminated with a two-sided valve and that in the spring they become more cylindrical, ending with a three-sided valve. Our observations are that after hibernation, the cases are still flat with a two-sided valve. Before pupation they turn more cylindrical with a three-sided valve.

Our findings indicate that Coleophora albicostella is rare in The Netherlands, even more rare than it appeared before we started working on this manuscript, and we have had to conclude that some of the records in the national database “NOCTUA” (Ellis 2015a) involved different species. In some cases this was due to identifications based on larval cases, but sometimes also because of a confusion with the similarly named C. albicosta (Haworth, 1828), which feeds on Ulex spp. (Fabaceae). Only a handful of verified C. albicostella records remain, all from the southernmost part of the country, the south of the province of Limburg. Outside this area, two doubtful records are reported, one from the southwest and one from the east of the country. C. gryphipennella on the other hand is abundant in The Netherlands with records throughout the country in all provinces (Küchlein and Donner 1993; Muus 2015). Most records are reported from the coastal area where Rosa pimpinellifolia is common, which is also reported to be the main host plant (Hering 1957).

Spatulate cases on Fragaria in the spring or the larger cases in the autumn are most likely all Coleophora gryphipennella, although we cannot completely exclude C. albicostella. It appears that the final cases in the spring are trivalved (see Fig. 11) and from this stage the adults emerge about a month later (Fig. 15). Small pistol cases (i.e. <5 mm) that may be encountered in the autumn on Fragaria are C. lutipennella or C. flavipennella, and can most likely be found in the vicinity of oaks (Quercus spp.), the common host for these species (Hering 1957). C. lutipennella and C. flavipennella are in the same species group as C. gryphipennella (sensuEmmet et al. 1996), but we find it unlikely that they can complete their larval stage without oaks. It is more likely that, when they come down from the oaks in the autumn to find a safe place to hibernate, they consume small amounts of Fragaria opportunistically.